starlight-spec
Skill1 starsChanged 4 months agoMITView on GitHub
Convert an approved product intent into a backlog-ready specification with boundaries, contracts, tasks, verification, rollback, and human approval points. Use when requirements are ambiguous, multiple components must change, or implementation needs a durable handoff.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- Starlight-Intelligence/starlight-agent-skillsword for word, first seen Nov 8, 2026
Other versions1 project uses a version of this.
Word for word · 1
- Starlight-Intelligence/starlight-agent-skills★ 1100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository43 other files it gives its agents.
- AGENTS.mdAGENTS.md
- adapters/claude.mdCLAUDE.md
- arcanea-mythic-overlaySkill
- frankx-authority-postSkill
- starlight-voiceSkill
- cosmic-code-labSkill
- starlight-autoplanSkill
- starlight-ceo-reviewSkill
- starlight-design-reviewSkill
- starlight-devex-reviewSkill
- starlight-eng-reviewSkill
- starlight-investigateSkill
- starlight-qaSkill
- starlight-shipSkill
- apod-to-shortSkill
- arxiv-space-paper-to-briefSkill
- cosmic-mythic-overlaySkill
- nasa-image-to-atlas-pageSkill
- rights-check-nasa-esaSkill
- rocket-launch-to-reelSkill
- space-social-repurposerSkill
- coding-challenge-generatorSkill
- explain-like-cosmic-professorSkill
- simulation-lab-builderSkill
- caption-and-hashtagSkill
- nasa-image-to-shortSkill
- rocket-launch-to-youtubeSkill
- social-repurposerSkill
- thumbnail-conceptSkill
- arxiv-paper-to-briefSkill
- claim-verificationSkill
- mission-page-to-summarySkill
- media-studioSkill
- publishing-studioSkill
- software-studioSkill
- template-studioSkill
- world-experience-studioSkill
- affiliate-auditSkill
- agentic-incomeSkill
- capability-foundrySkill
- notion-operating-systemSkill
- payments-mandateSkill
- swarm-queen-coordinationSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 11
Create an executable specification before code when the cost of ambiguity is greater than the cost of writing the contract.
- Distinctive: line 52
Read the repository and its local instructions before naming files or commands.
- Distinctive: line 59
No invented APIs, versions, metrics, or approvals.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Starlight Spec
Purpose
Create an executable specification before code when the cost of ambiguity is greater than the cost of writing the contract.
When it fires
- A feature spans more than one component or repository.
- Inputs, outputs, states, permissions, or failure behavior are unclear.
- The user asks for a PRD, technical design, or implementation plan.
Inputs
- User outcome, audience, and decision owner.
- Existing architecture, conventions, and repository instructions.
- In-scope and out-of-scope behavior.
- Data sensitivity, runtime targets, compatibility, and delivery constraints.
Workflow
- Define the user-visible outcome and measurable acceptance criteria.
- Record non-goals, assumptions, unknowns, and authority boundaries.
- Map existing components and identify the smallest required change surface.
- Specify data shapes, state transitions, APIs, errors, and compatibility behavior.
- Decompose work by dependency and verification boundary, not an arbitrary line count.
- Define tests, preview checks, observability, rollback, migration, and promotion gates.
- Mark any unresolved choice that would materially change the design for human decision.
Output contract
Return a specification containing:
- problem and user outcome;
- scope and non-goals;
- current-state evidence;
- functional and non-functional requirements;
- data, API, state, trust, and compatibility contracts;
- ordered implementation tasks with ownership;
- acceptance tests and manual verification;
- rollout, rollback, and migration plan;
- open decisions and human gates.
Tools & MCP
Read the repository and its local instructions before naming files or commands. Browse primary documentation when an external API or current standard is involved. Use read-only inspection by default.
Quality bar
- File and symbol references exist.
- Requirements are testable and do not smuggle in a preferred implementation.
- Failure, empty, loading, offline, and permission states are considered where relevant.
- No invented APIs, versions, metrics, or approvals.
- A different implementer can execute the spec without guessing the intent.
Example
Input: “Let customers install one department on Codex and Claude.”
Good output: separate package, host adapter, installation, verification, rollback, and publication contracts, with marketplace approval and credentials explicitly outside autonomous authority.
Built on SIP — Starlight Intelligence Protocol Vertical: starlight-agent-skills · portable capability layer
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

